Richard L. Phillips   

@rlanasphillips

Research in theory and ML. Interpretability, fairness, and responsible prediction. Amateur cricket player and chemist. He/him

Ithaca, NY
Vrijeme pridruživanja: ožujak 2017.

Tweetovi

Blokirali ste korisnika/cu @rlanasphillips

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @rlanasphillips

  1. proslijedio/la je Tweet
    1. velj
    Poništi
  2. proslijedio/la je Tweet

    Best CS paper: Kate Donahue and Jon Kleinberg, Fairness and Utilization in Allocating Resources with Uncertain Demand 2/

    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    30. sij

    The recent work just mentioned on how explanations can be gamed is by et al:

    Poništi
  4. proslijedio/la je Tweet
    28. sij

    This fairness portability problem discuss is very real. One version of the Idaho pretrial risk assessment bill required validation, but did not specify that they should be validated *in Idaho*.

    Poništi
  5. proslijedio/la je Tweet
    6. sij

    Most popular computer science paper of the day: "How can we fool LIME and SHAP? Adversarial Attacks on Post hoc Explanation Methods"

    Poništi
  6. proslijedio/la je Tweet
    3. sij

    It’s grad application season and I see a lot of “I would never take a student who didn’t contact me!” comments. A reminder that there are a lot of first-gen students out there who might not know that this is a thing and they deserve a shot at discussing their interests w/ you.

    Prikaži ovu nit
    Poništi
  7. proslijedio/la je Tweet
    30. pro 2019.

    I've seen this a bunch in my timeline yesterday and today. As the person running Cornell CS PhD admissions this year, I have Thoughts (some of which have been stewing for longer than the past day).

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    28. pro 2019.

    I'm sure everyone saw this live on C-SPAN 2 just now, so there is hardly a need to point it out. But in case a few of you were watching C-SPAN 1 instead, check on and me talking about the Ethical Algorithm on BookTV:

    Poništi
  9. proslijedio/la je Tweet
    20. pro 2019.

    counterfactuals, what would we do without them

    Prikaži ovu nit
    Poništi
  10. proslijedio/la je Tweet
    17. pro 2019.

    If deep learning is summoning demons, interpretability is demon biology. 🔬👹

    Poništi
  11. proslijedio/la je Tweet
    Poništi
  12. proslijedio/la je Tweet

    When I started my PhD, I wasn't sure I'd be able to finish, let alone have my graduation be featured in the BBC, Atl Black Star, & my uni Here's to Black CS PhDs not being a rarity & the Black women who empower me

    Poništi
  13. proslijedio/la je Tweet
    14. pro 2019.

    But *why* do we try to be so fine-grained? Because we really REALLY dislike the idea of someone getting more than they "deserve" from public programs. We have an extremely high cost function for "getting undue help" and we don't measure the (high) cost of being so fine-grained.

    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    13. pro 2019.

    Graph representation learning is the most popular workshop of the day at . Amazing how far the field has advanced. I did not imagine so many people would get into this when I started working on graph neural nets back in 2015 during an internship. Time flies...

    Poništi
  15. proslijedio/la je Tweet
    13. pro 2019.

    Black scholars from 42 countries were able to come present technical work and attend (!) but rejected visas still interfere with bringing outstanding scholars here. If the Canadian government is serious about supporting AI, fixing this broken process is how.

    Poništi
  16. proslijedio/la je Tweet
    13. pro 2019.
    Poništi
  17. proslijedio/la je Tweet
    Poništi
  18. proslijedio/la je Tweet
    12. pro 2019.

    Check out @CharlieMarx9 and at NeurIPS poster #107! Feature correlation is confusing in influence methods and they have answers!

    Poništi
  19. proslijedio/la je Tweet
    12. pro 2019.

    Really nice to see all the interest in talking to @CharlieMarx9 and about our work on indirect feature influence today at ! Paper:

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·